Without limiting the invention, its background is described in connection with a process for the manufacture of sub-100 nm devices using imprint lithography.
In manufacturing, lithography techniques that are used for large-scale production include photolithography and other application oriented lithography techniques, such as electron beam lithography, ion-beam and x-ray lithography, as examples. Imprint lithography is a type of lithography that differs from these techniques. Recent research has shown that imprint lithography techniques can print features that are smaller than 50 nm. As such, imprint lithography has the potential to replace photolithography as the choice for semiconductor manufacturing in the sub-100 nm regime. It can also enable cost effective manufacturing of various kinds of devices, including patterned magnetic media for data storage, micro optical devices, MEMS, biological and chemical devices, X-ray optical devices, etc.
Current research in the area of imprint lithography has revealed a need for devices that can perform orientation alignment motions between a template, which contains the imprint image, and a substrate, which receives the image. Of critical importance is the careful and precise control of the gap between the template and the substrate. To be successful, the gap may need to be controlled within a few nanometers across the imprinting area, while, at the same time, relative lateral motions between the template and the substrate must be eliminated. This absence of relative motion leads is also preferred since it allows for a complete separation of the gap control problem from the overlay alignment problem.
For the specific purpose of imprinting, it is necessary to maintain two flat surfaces as close to each other as possible and nearly parallel. This requirement is very stringent as compared to other proximity lithography techniques. Specifically, an average gap of about 100 nm with a variation of less than 50 nm across the imprinted area is required for the imprint process to be successful at sub-100 nm scales. For features that are larger, such as, for example, MEMS or micro optical devices, the requirement is less stringent. Since imprint processes inevitably involve forces between the template and the wafer, it is also desirable to maintain the wafer surface as stationary as possible during imprinting and separation processes. Overlay alignment is required to accurately align two adjacent layers of a device that includes multiple lithographically fabricated layers. Wafer motion in the x-y plane can cause loss of registration for overlay alignment.
Prior art references related to orientation and motion control include U.S. Pat. No. 4,098,001, entitled “Remote Center Compliance System;” U.S. Pat. No. 4,202,107, entitled “Remote Axis Admittance System,” both by Paul C. Watson; and U.S. Pat. No. 4,355,469 entitled “Folded Remote Center Compliant Device” by James L. Nevins and Joseph Padavano. These patents relate to fine decoupled orientation stages suitable for aiding insertion and mating maneuvers in robotic machines and docking and assembly equipment. The similarity between these prior art patents and the present invention is in the provision for deformable components that generate rotational motion about a remote center. Such rotational motion is generated, for example, via deformations of three cylindrical components that connect an operator and a subject in parallel.
The prior art patents do not, however, disclose designs with the necessary high stiffness to avoid lateral and twisting motions. In fact, such lateral motion is desirable in automated assembly to overcome mis-alignments during the assembly process. Such motion is highly undesirable in imprint lithography since it leads to unwanted overlay errors and could lead to shearing of fabricated structures. Therefore, the kinematic requirements of automated assembly are distinct from the requirements of high precision imprint lithography. The design shown in U.S. Pat. No. 4,355,469 is intended to accommodate larger lateral and rotational error than the designs shown in the first two patents, but this design does not have the capability to constrain undesirable lateral and twisting motions for imprint lithography.
Another prior art method is disclosed in U.S. Pat. No. 5,772,905 (the '905 patent) by Stephen Y. Chou, which describes a lithographic method and apparatus for creating ultra-fine (sub-25 nm) patterns in a thin film coated on a substrate in which a mold having at least one protruding feature is pressed into a thin film carried on a substrate. The protruding feature in the mold creates a recess of the thin film. First, the mold is removed from the film. The thin film is then processed such that the thin film in the recess is removed exposing the underlying substrate. Thus, the patterns in the mold are replaced in the thin film, completing the lithography. The patterns in the thin film will be, in subsequent processes, reproduced in the substrate or in another material which is added onto the substrate.
The process of the '905 patent involves the use of high pressures and high temperatures to emboss features on a material using micro molding. The use of high temperatures and pressures, however, is undesirable in imprint lithography since they result in unwanted stresses being placed on the device. For example, high temperatures cause variations in the expansion of the template and the substrate. Since the template and the substrate are often made of different materials, expansion creates serious layer-to-layer alignment problems. To avoid differences in expansion, the same material can be used but this limits material choices and increases overall costs of fabrication. Ideally, imprint lithography could be carried out at room temperatures and low pressures.
Moreover, the '905 patent provides no details relative to the actual apparatus or equipment that would be used to achieve the process. In order to implement any imprint lithography process in a production setting, a carefully designed system must be utilized. Thus, a machine that can provide robust operation in a production setting is required. The '905 patent does not teach, suggest or disclose such a system or a machine.
Another issue relates to separation of the template from the substrate following imprinting. Typically, due to the nearly uniform contact area at the template-to-substrate interface, a large separation force is needed to pull the layers apart. Such force, however, could lead to shearing and/or destruction of the features imprinted on the substrate, resulting in decreased yields.
In short, currently available orientation and overlay alignment methods are unsuitable for use with imprint lithography. A coupling between desirable orientation alignment and undesirable lateral motions can lead to repeated costly overlay alignment errors whenever orientation adjustments are required prior to printing of a field (a field could be for example a 1″ by 1″ region of an 8″ wafer).
Further development of precise stages for robust implementation of imprint lithography is required for large-scale imprint lithography manufacturing. As such, a need exists for an improved imprint lithography process. A way of using imprint lithography as a fabrication technique without high pressures and high temperatures would provide numerous advantages.